ROBOT – Layoff Range Confusion?

[HorseProfiler]

Your help is greatly appreciated!

For example tomorrow DMR R#6, I’m having great difficulty utilizing the layoff range 2nd and 3rd. Layoff 1rst works perfect, by setting range to 2 – 380 it returns 6 bets and by setting range to 2 – 377 the horse coming off the 378 layoff drops from the list which now contains 5 bets.

For Layoff 2nd all these range settings return “0 bets”: 2 – 400, 2 – 399, 49 – 400, 2 – 90 default. It’s important that I can accurately utilize this feature, please help. Below are the layoff dates for each horse, read across:

#1 22 49 77
#2 378 400 428
#3 24 49 120
#4 91 120 295
#5 45 92 115
#6 77 91 120

[Rick]

This is from the Robot User Guide:

2nd After Layoff
The second start after a layoff can be a productive angle. Try various day ranges that would produce a
query such as “2nd after a layoff of 90 days or more”. Here is how you would set that up

[ x ] Layoff Range
[ * ] 2nd
Min Days (left) 090
Max Days (right) set to 990 (maximum)

The results of this test would be horses that had run last out after a layoff of 90-days or more and
now are returning to race again (2nd after layoff).

Additional notes on the ‘Layoff’ box.. The default setting is “1st” (days since raced).

[x] Layoff Range Activates the layoff filter and all FTS are ignored.
{*)1st This is the typical layoff (LAY) days found throughout HTR2 – ‘days since raced’
(*)2nd Second start after layoff of xxx days.
(*)3rd Third start after a layoff of xxx days.

[Rick]

Try setting minimum to at least 45.

[km]

Liam, you may confused on the day logistics

Let's look at each of your examples

#1 22 49 77
#2 378 400 428
#3 24 49 120
#4 91 120 295
#5 45 92 115
#6 77 91 120

1) 1st after layoff 22 days; 2nd after layoff 27 days, 3rd after layoff 29 days

2) 1st after layoff 378 days; 2nd after layoff 22 days, 3rd after layoff 28 days

3) 1st after layoff of 24 days, 2nd after layoff 25 days, 3rd after layoff of 71 days

4) 1st after layoff of 91 days, 2nd after a layoff of 29 days, 3rd after layoff 175 days

etc.

count the days in between to determine the 2nd and 3rd.

(4) is the most interesting of the above. Although it is a 3rd after layoff of 175 days, it also has had another layoff of 90 days since its last start, perhaps the negating the value of the "3rd after layoff".

The Robot is not sophisticated enough to allow multiple instances of layoff. For example, you cannot eliminate #4 with "3rd after layoff of 90 or more" IF "1st after layoff < 30 days"

You'll need a db query to handle that, using HX5.

But i'll caution you before you enter this quagmire of layoff days parameter testing, there is little to pin your ROI on for "2nd" and "3rd" The concentration should be with 1st after layoff 029 - 165, also known as a "freshening". Trainers are rarely mapping out their horse's race spacing with "2nd" and "3rd" after layoff, it just happens according to condition book, track scheduling, injuries, etc., but they do purposely give short vacations after the last start and that's where the ROI can be found.

[DanG]

But i'll caution you before you enter this quagmire of layoff days parameter testing, there is little to pin your ROI on for "2nd" and "3rd" The concentration should be with 1st after layoff 029 - 165, also known as a "freshening". Trainers are rarely mapping out their horse's race spacing with "2nd" and "3rd" after layoff, it just happens according to condition book, track scheduling, injuries, etc., but they do purposely give short vacations after the last start and that's where the ROI can be found.
Good point Ken; the 2nd and 3rd off a layoff (I thought) was going to yield some great profits. Experience and anecdotal evidence suggested it was a powerful angle, but large sample studies have produced as many misses as hits with that angle. Even trainers who have shown some past success (or failure) have been spotty going forward.

As always; if it works for the individual player; then it WORKS period. But; it’s been something I’ve largely ignored unless it’s clearly a case of sprinting preps for a route, or a turf conditioner etc.

[Huguenot]

So Ken, if I use 2nd or 3rd off a layoff, does the ROBOT give me a reading on EVERY horse that fits this description OR does it only return plays on horses who have been off less than 30 days.
In other words,

23 23 175 (YES)
40 23 175 (NO)

23 50 175 ?????? just 2nd off a layoff? or both 2nd and 3rd off a layoff

[Rick]

Remember, you need to set the minimum range of what you are looking for in a 2nd or 3rd start after layoff.

Good place to start is 45 days. Then you can compare the results you get to the FPS screen. That screen shows an underscore after the date for a layoff of 45 days or more.

[km]

As i stated in my previous thread Andy, the Robot does not perform multiple instances of layoff.

If you set it for "3rd after layoff" and 180 - 999 days, then the following horses would be picked up =
(number of days back for each of the last 3 starts)

a) 30, 90, 300
b) 90, 180, 365
c) 180, 360, 540

In case (a), the time between the horse's 3rd start back and next start was 210 days. So today's start is "3rd after layoff of 180 or more". The 30 days since last raced is not considered.

(b) the time between the 3rd back and 2nd back was 185, so today's start is "3rd after a layoff of 180+" even though he has had another layoff of 90 days since his last, but that is not considered by the "3rd after layoff" filter.

(c) this horse has multiple long layoff gaps but still fufills the criteria of "3rd after layoff 180+"

I balk at putting in multiple-level-filtering of layoff days into Robot2 without causing a complex nightmare of statistical illogic that would confuse us all. Even db queries that attempt something like =

"3rd after layoff 180+" AND "2nd After layoff 30 or less" AND "last start (1st after layoff) less then 30" is very complicated to code and difficult to verify because it will always lead to exceptions, such as a horse with just two lifetime starts sneaking into the mix.

My advice is the same as Dan's = get the "3rd after layoff" list if you wish and eyeball the horses and consider the trainer stats. But don't get into deep with all the permutations on layoffs, the reward does not match the brain energy invested.

[HorseProfiler]

Ken, Rick and Dan thanks for all your input. Ken your analysis confirms my query is working perfect and our numbers match; my confusion is regarding why my db query results don't match Robots.

Utilizing the example race (FLD=6), the results of my db query and Robot don’t match. *Identify horses making 3rd start after a “freshening” of 29 - 165

Layoff Range 3rd Robot – (29 – 165) 1 bet (#3) My db 2 bets #3 - #6
Layoff Range 3rd Robot – (29 – 175) 1 bet (#3) My db 3 bets #3 - #4 - #6

The answer probably is as you say Ken; Robot does not allow multiple instances of layoffs and one may cancel out the other. By utilizing the HX5 nPPxDATE's fields my DateDiff query simply looks down the Layoff 3rd column and assigns the appropriate value regardless of the Layoff 2nd column and Layoff 1rst column.